NVIDIA BlueField BSP v4.8.0
NVIDIA BlueField BSP v4.8.0 Download PDF

Release Notes Change Log History

  • Added support for new BlueField reset and reboot procedures for loading new firmware and firmware configuration changes which replace previous need for server power cycle

  • Updated the default operation mode of SuperNICs to NIC mode (from DPU mode). This is relevant to the following SKUs:

    • 900-9D3B4-00CC-EA0

    • 900-9D3B4-00SC-EA0

    • 900-9D3B4-00CV-EA0

    • 900-9D3B4-00SV-EA0

    • 900-9D3B4-00EN-EA0

    • 900-9D3B4-00PN-EA0

    • 900-9D3D4-00EN-HA0

    • 900-9D3D4-00NN-HA0

      Note

      When upgrading one of these SuperNICs to 2.7.0, if its mode of operation was changed at any point in the past, then the last configured mode of operation will remain unchanged. Otherwise, the SuperNIC will rise in NIC operation mode.

  • Installing the BFB Bundle now performs NIC firmware update by default

  • Added ability to install NIC firmware and BMC software in NIC mode in NVIDIA® BlueField®-3.

    Note

    It is important to note the following:

    • During BFB Bundle installation, Linux is expected to boot to upgrade NIC firmware and BMC software

    • As Linux is booting during BFB Bundle installation, it is expected for the mlx5 core driver to timeout on the BlueField Arm

    • During the BFB Bundle installation, it is expected for the mlx5 driver to error messages on the x86 host. These prints may be ignored as they are resolved by a mandatory, post-installation power cycle.

    • It is mandatory to power cycle the host after the installation is complete for the changes to take effect

  • Software packaging – new BlueField firmware bundle package (bf-fwbundle-<version>.prod.bfb), a smaller image for Day 2 upgrades, without the OS and DOCA runtime. Includes ATF, UEFI, nic-fw, bmc-fw, and eROT only.

  • Improved BlueField BMC robustness –

    • Report LLDP for L2 discovery via Redfish

    • Improved BlueField DPU debuggability

  • Increased support for virtio-net VF devices on BlueField-3 networking platforms to 2K

  • Reduced power consumption for BlueField NIC mode

  • RAS

    • Report DDR Error to OS, including both single-bit ECC error and UCE error

    • Support error injection in processors, memory, and PCIe devices

  • Updated minimum UEFI password requirements

  • Included DPU BMC firmware as part of the BFB image

  • Added virtio-net support for plugging/unplugging parallel devices

  • Implemented virtio debug enhancements

  • Added Redfish support for configuring all UEFI secure boot settings (disable, enable, enroll user keys, etc.) at scale, remotely, and securely

  • For FHHL DPUs, added support for performing PCIe bifurcation configuration via MFT tool

    Note

    Only a subset of configurations are supported.

  • Updated the print of the manufacturing (MFG) setting, MFG_OOB_MAC, displayed by the command bfcfg -d to appear in lower-case to align with standard Linux tools

Note

Upgrading to this BSP version installs a new version of Ubuntu GRUB. This version of GRUB revokes the old UEFI secure boot certificates and install new ones. The new certificates will not validate older images and boot will fail. Therefore, to roll back to older software versions, users must disable UEFI secure boot.

  • BFB installation chooses the on-chip NVMe (/dev/nvme0n1) by default for the EFI system partition and Linux rootfs installation and can be overloaded with device=/dev/mmcblk0 in bf.cfg to push together with the BFB.

    Info

    Installing on NVMe causes DPU booting to stay at the UEFI shell when changing to Livefish mode.

    Info

    A previously installed OS on the eMMC device stays intact. Only the EFI boot entry is updated to boot from the SSD device.

  • BlueField-3 tuning update for power and performance

  • BlueField-3 power-capping and thermal-throttling

  • Added Linux fsck to boot flow

  • Log PCIe errors (to RShim log)

  • Halt uncorrectable double-bit ECC error on DDR

  • Added support for live migration of VirtIO-net and VirtIO-blk VFs from one VM to another. Requires working with the new vDPA driver.

  • OS configuration – enabled tmpfs in /tmp

  • Added support for Arm host

  • Enroll new NVIDIA certificates to DPU UEFI database

    Warning

    Important: User action required! See known issue #3077361 for details.

Note

This is the last release to offer GA support for first-generation NVIDIA® BlueField® DPUs.

  • Added support for NIC mode of operation

  • Added password protection to change boot parameters in GRUB menu

  • Added IB support for DOCA runtime and dev environment

  • Implemented RShim PF interrupts

  • Virtio-net-controller is split to 2 processes for fast recovery after service restart

  • Added support for live virtio-net controller upgrade instead of performing a full restart

  • Expanded BlueField-2 PCIe bus number range to 254 (0-253)

  • Added a new CAP field, log_max_queue_depth (value can be set to 2K/4K), to indicate the maximal NVMe SQ and CQ sizes supported by firmware. This can be used by NVMe controllers or by non-NVMe drivers which do not rely on NVMe CAP field.

  • Added ability for the RShim driver to still work when the host is in secure boot mode

  • Added bfb-info command which provides the breakdown of the software components bundled in the BFB package

  • Added support for rate limiting VF groups

  • PXE boot option is enabled automatically and is available for the ConnectX and OOB network interfaces

  • Added Vendor Class option "BF2Client" in DHCP request for PXE boot to identify card

  • Updated the "force PXE" functionality to continue to retry PXE boot entries until successful. A configuration called "boot override retry" has been added. With this configured, UEFI does not rebuild the boot entries after all boot options are attempted but loops through the PXE boot options until booting is successful. Once successful, the boot override entry configuration is disabled and would need to be reenabled for future boots.

  • Added ability to change the CPU clock dynamically according to the temperature and other sensors of the DPU. If the power consumption reaches close to the maximum allowed, the software module decreases the CPU clock rate to ensure that the power consumption does not cross the system limit.

    Note

    This feature is relevant only for OPNs MBF2H516C-CESOT, MBF2M516C-EECOT, MBF2H516C-EESOT, and MBF2H516C-CECOT.

  • Bug fixes

  • Added ability to perform warm reboot on BlueField-2 based devices

  • Added support for DPU BMC with OpenBMC​

  • Added support for NVIDIA Converged Accelerator (900-21004-0030-000)

© Copyright 2024, NVIDIA. Last updated on Aug 20, 2024.